Search code examples
splitnumbersocrtesseractspacing

How to separate (space between) numbers in tesseract OCR

I try to get numbers from image

but after submitting my result is 2 332223355 1 23, i don't really understand how does it splits, everything i need is to split one, two and three digit numbers with space. can anybody help me?

Solution

  • Use:

    tesseract -psm 7 NXect.png stdout

    which gives for the image you provided: 

    2 3 32 22 33 55 123‘

    The tesseract version I am using: 

    $ tesseract --version
tesseract 3.04.01
 leptonica-1.73
  libgif 5.1.2 : libjpeg 8d (libjpeg-turbo 1.4.2) : libpng 1.2.54 : libtiff 4.0.6 : zlib 1.2.8 : libwebp 0.4.4 : libopenjp2 2.1.0

    gives me for the original image without any options: 

    Error in pixGenHalftoneMask: pix too small: w = 250, h = 58
23 32 22 33 55 123

    and for the resized image (2x): 

    $ tesseract  NXect_x2.png stdout
23 32 22 33 55 123

    so I can't confirm the OCR result you are getting out the image.